config.py centralizes phase-2 path/env reads (no hard-coded data paths,
section 15). prompt.py loads the EXTERNAL, versioned prompt pack (section
14): pack version + per-prompt content hash feed prompt_version for the
response-cache composite key (D6); missing/invalid packs raise a clear
PromptPackError, never a silent fallback. Minimal fixture pack under
tests/fixtures/prompt_pack/ pins CI to a local, versioned pack.
run_encounter gains llm_policy and per-run policy_factory (mutually
exclusive with policy/llm_policy); replay.py logs every (observation,
action) keyed by call_seq to JSONL and replay() reproduces the recorded
battle bit-identically (wins/draws/rounds/transcript) with mismatch,
exhausted, duplicate and out-of-order guards. P0 frontier complete: the
recorded run replays identically.
Add replay.py: a ReplayLog records every (observation, action) pair a
policy produced, keyed by monotonic call_seq, plus the EncounterSpec,
seed and round_cap, and serializes the whole battle to JSONL. replay()
turns a log back into an observation-only LLMPolicy that returns each
recorded action, raising ReplayMismatchError on a differing observation
and ReplayExhaustedError past the last call.
Observations are matched by their canonical JSON (observation_key), not
raw dataclass equality, because JSON collapses tuples and lists and the
snapshot's inner dict[str, object] holds both. Action and the spec are
dumped through Pydantic TypeAdapters (mode=json) so tuple fields survive
as arrays and are coerced back on load.
Move TrickCondition/ReadyTrigger out of the TYPE_CHECKING block in
combat.py: TypeAdapter(Action) must resolve them at runtime to serialize
the dataclass, so the Literal aliases now live at module level.
Compact render is 15,691 bytes (~3.9k tokens) = 24.4% of the raw 64,204-byte
to_dict() snapshot on the representative 7-combatant encounter, hitting the
~2-4k token target; grids are sparse and determinism is pinned in tests.
candidates() derives plausible actions from the pairs with no second legality
engine (swift/free/full_round excluded).
Build a decision-sufficient ObservationView from the raw snapshot and
render it as a deterministic line-based prompt string. The raw to_dict()
is ~64 KB (~16k tokens) for a 7-combatant encounter; the compact view
renders at ~15.7 KB (~3.9k tokens, 24% of raw), under the 4k-token
ceiling of the phase-2 architecture doc.
Compaction: sparse terrain/elevation/threatened grids, per-combatant
key selection (occupied_cells, abilities, skills, initiative_mod/total,
temp_hp_rounds_left, features, carried_items, weapon_damage, aware,
alignment, armor_spell_failure, arcane_caster, run_rounds dropped),
attacks compacted to formula+type damage strings, spells to 7
decision-relevant fields, pair target_index replaced by the stable
combatant id, condition_definitions dropped. Attacks and spells render
as positional tuples declared by attack_fields/spell_fields legend
lines, with '-' for missing values.
Tests pin the size bounds (compact < 0.25 x raw and < 16000 bytes), the
sparse-grid contract, the dropped-key contract and the deterministic
id/render contract on a caster-heavy 3v4 representative battle.
Measures: observation 64 KB (~16k tokens) for 7 combatants / 22 KB for a 2v2;
124 policy calls per battle, i.e. 124k calls per 1000-run sweep cell -> a
per-turn LLM in a sweep is infeasible (D9 validated) and a compact view is
mandatory. Danger map (default vs focus-fire monsters, 100 runs): the default
policy understates danger (PC deaths 13%->24% and 0%->19%, TPK 0%->2%) and the
win rate alone misleads (rises to 98% while deaths rise) -> D12/D13 validated.
Notes the spike limits and the remaining P0.2-P0.5 items.
run_encounter gains an opt-in collect_runs flag returning one RunOutcome
per battle (winner, rounds, per-combatant id/side/hp/hp_max/dead) next to
the EncounterReport; default behavior is unchanged. objectives.py provides
the pure optimizer primitives: P(survives all), P(TPK), P(>=1 death),
expected hp lost, expected rounds and a worst-decile (CVaR-style) tail
metric. phase2_spike.py holds the spike-level strengthened monster policy
(focus fire on the lowest-AC enemy, ties to the wounded; party side stays
on default_policy). tests/test_phase2_danger_map.py prints the danger-map
sketch: warband and lich scenarios measured default vs focus-fire at
100 runs / seed 11, pinning the delta deterministically.
Representative 7-combatant caster-heavy encounter on sample_arena,
measured through the real engine path (engine._build_observation at each
policy call, delegating to default_policy):
- observation: 64,204 bytes at first call, 67,712 max mid-battle (~16.9k tokens)
- policy calls/battle: 124 (7 combatants, round_cap=20), 12 for a 2v2 duel
- projected calls per sweep cell: 124 / 12,400 / 124,000 at runs 1/100/1000
Loose bounds (obs < 200 KB, calls/battle < 200) + seed-determinism pinned.
- LLMPolicy is now Callable[[Observation], tuple[Action, ...]]: the snapshot
is the policy's sole information source; the live CombatantState is no
longer passed to new-style policies.
- CombatEngine._llm_adapter keeps a compatibility path for pre-D1
two-argument callables: the signature is inspected ONCE at wrap time
(_accepts_state) and legacy policies still receive (observation, state).
- Observation.to_dict() carries schema_version (OBSERVATION_SCHEMA_VERSION=1)
as its first pinned key; bump only on removal/meaning change.
- Pinned keyset tests updated (test_observation, test_observation_enriched);
new tests: observation-only e2e, legacy 2-arg e2e through the adapter,
schema_version pin + JSON-serializability.
- README: documents the new entry signature and the schema_version key.
All recommendations accepted by the user except prompt location: prompts live
outside this repo (separate pack, configurable path, versioned; prompt_version
feeds the cache key and the replay record; a pinned pack for integration, a
minimal local fixture for unit tests, no silent fallback). Also records the
repo-scope decision (code only: calculation + LLM + report) with the final
detachment of data/sheets/prompts as a backlog item and the design constraint
to avoid hardcoded data paths in the new code. Priority: cartography first.
The execution tiers answer 'how to run a tactic', not 'which tactic is best'
nor 'where is the danger'. Adds two decision layers: an optimizer over
StrategySpec (objectives incl. tail risk, coarse-to-fine search, LLM as
candidate seeder) and an adversarial cartography (AdversarySpec + alternating
min-max + scenario sweeps + danger frontier + sensitivity), giving the danger
view for BOTH sides. Reframes the LLM as hypothesis generator / red team /
explicator / reporter, never the optimizer or per-turn executor in sweeps.
Adds D12-D14 and phases P3 (optimizer) / P4 (robustness).
Key correction: the LLM must not sit on the per-turn hot path of a Monte
Carlo sweep (runs x rounds x combatants calls, no cross-run cache reuse).
Adopts a two-tier design - a compiled, parameterised deterministic Policy for
quantitative sweeps (tier A) and a per-turn NLPolicy for a handful of showcase
battles (tier B). Also fixes: 'engine unchanged' -> rules unchanged/additive
boundary; replay keyed by a monotonic call_seq (delay re-queues, summons);
composite response-cache key incl. model id; ObservationView as a structured
object; candidates derived from pairs instead of a second legality engine;
per-run PolicyFactory; cost guardrails in P1; concurrency; D1 to settle in P0.
Design-only document (no code): target architecture (compact versioned
observation view, legal-actions API, LLM client + cache, NL policy with
parse/validate/repair/fallback, strategy sweep, replay/eval harness), the
concrete gaps in the current runner/CLI wiring, the decisions to settle and a
phased implementation plan (P0 foundations, P1 client+policy, P2 strategy
sweep, P3 provider+report) with exit criteria and risks.
Brings in the last four out-of-scope mechanics: the nonlethal damage track
(unarmed strikes, minimum damage), splash weapons (direct + scatter),
combat skills (skill checks, Acrobatics Tumbling, Stealth/opposed Perception)
and 3D flight over ground terrain (vertical line of effect deferred). Updates
the out-of-scope and residual tables to drop the now-modeled items and marks
the round-2 disposition tables as historical. Gate 1981.
A read-only audit of the transcript rendering, map and loader formats found
14 confirmed bugs (4 Med, 10 Low: duplicate --side, id-disambiguation
collision, range_increment_ft division by zero, silent spell_slots/features
drops, negative round-cap, multi-char legend keys, blank rows, id parsing,
duplicate YAML keys, move_cost default) - all fixed. The transcript/format
area is audited; only the structural out-of-scope items remain. Gate 1900.
The README documents deployment as required when zones exist, but load_map
only enforced the reverse direction (deployment without zones). A map with
a zones layer and no deployment now fails with MapValidationError instead
of loading with unreachable deployment zones.
_merge_resource_grants silently dropped non-int pool values when features
were present (a "5" pool vanished, leaving only the feat grant) while
Pydantic coerced the same value when features were absent. _validated_resources
now runs the JSON resources through TypeAdapter(dict[str, int]) -- the same
lax coercion as Combatant.resources -- so both paths coerce-or-fail
identically, and grants are added to the coerced ints.
Feat resolution read level with an isinstance(int) check and fell back
to 1, while Pydantic accepts numeric strings and lossless floats: a
level-4 sheet authored as "4" got Toughness for level 1. _parse_level
now runs the value through TypeAdapter(int) (same lax coercion as the
model) and raises CombatantLoadError for values the model would reject.
features was popped and only re-added when it was a list, so a bare
string (or dict/null) loaded with zero feats - silent data loss. Now a
present non-list features raises CombatantLoadError, and the list
resolution adopts an accept-string-or-error contract: non-string or
blank entries raise instead of being silently dropped.
int(k) on a non-numeric JSON key escaped the loader as an unhandled
ValueError traceback (CLI exit 1) instead of the documented exit code 2.
Route the conversion through _parse_spell_slot_keys which wraps it in a
CombatantLoadError with a clear message. Audited the other loader-side
int()/conversions: foundry.py guards every int() with isinstance/isdigit
checks and combatant.py has no other unguarded conversion.
Sunder now uses object hardness/HP with the SRD broken condition (usable at
-2), steal transfers real carried items and unarmed disarm auto-picks up the
dropped weapon. Objects/inventory leave the out-of-scope list; skills,
nonlethal, splash, 3D flight and the transcript remain. Gate 1861.
SRD Sunder + CRB 'Smashing an Object' / 'Conditions - Broken': weapons
carry Table 7-12 object stats (AttackSpec.hardness/hit_points/enhancement,
defaults by weapon class); a successful sunder deals the attacker's actual
physical weapon damage minus the target weapon's hardness, subtracted from
its hp. At <= half hp the weapon gains the broken condition - still usable
at -2 attack/damage, crits only on a natural 20 at x2. At <= 0 it is
destroyed (weapon_destroyed, unusable). The old flat str_mod+2 / durability
10 / synthetic 'disarmed' condition is gone. Firearm misfire carve-out kept
(broken firearms stay unusable until Quick Clear - misfire escalation not
modeled, documented).
107/108 spells carry their SRD components; arcane_caster and
armor_spell_failure are set from the sheets; the deafened 20% verbal
failure and the arcane spell failure roll are modeled. Components/ASF
leave the out-of-scope list. Gate 1850.
The components/ASF bullet no longer defers the data encoding (done in
the two preceding commits): 107/108 registry spells carry their KB
components (Create Alcool is a homebrew cantrip with no KB entry) and
all 15 combatants carry arcane_caster plus an effective
armor_spell_failure - zero everywhere because the shipped arcane casters
are unarmored or wear light armor whose ASF their magus/bard feature
annuls for their own-list spells. Also collapses an accidental
duplicated armor_spell_failure/arcane_caster description in the
observation snapshot bullet.
- arcane_caster true: misty (Magus/Bard sheet class), nairda (Witch/
Sorcerer), thrune_agent (Human magus 8, KB bestiary/thrune-agent.md);
lich already true. False: oni (Cleric), tammara (Gunslinger/
Warpriest - divine), esha (Cavalier/Rogue, spell-like only), the
companions harvie/ierlieth/jeanne/esha_mount and the remaining
monsters.
- armor_spell_failure stays 0 everywhere, as the effective PF1e value:
nairda and the lich wear no armor (sheet note "Pas d'armure", KB lich
block); misty's mithral shirt (RAW ASF 10%, magic-items/mithral-shirt
.md) and the thrune agent's +1 studded leather (15%) are light armor
and every spell they cast is on their own class list - magus and bard
cast their own spells in light armor with no arcane spell failure
(classes/magus.md, classes/bard.md).
- tests/test_spell_failure.py: the lich/oni smoke test becomes a
full-data table pinning all 15 combatants' flags with per-entry
sources. No shipped combatant incurs a non-zero ASF; the failure roll
itself stays covered by the synthetic armored-caster tests.
- Every data/spells/*.json now carries a "components" field with the
V/S/M/F/DF tokens from its KB Components line (material/focus
descriptions dropped; M/DF and F/DF alternatives kept verbatim).
- 107/108 encoded. Create Alcool (data/spells/create_alcool.json) is
left without the field: it is a homebrew cantrip (level 0) with no KB
entry - the closest real spell, Enhance Water (Inner Sea Gods, V, S),
is already a separate registry entry.
- summon_monster_i/ii map to the KB's numeric files (summon-monster-1/2);
hidden_diplomacy maps to its KB name Egorian Diplomacy (same spell,
Spymaster's Handbook rename): V, S, M.
- tests/test_spell_failure.py: the placeholder asserting the registry was
not yet encoded is replaced by an encoded-state smoke test (all spells
but Create Alcool declare components; five spot-checks against the SRD
stat blocks). Golden transcripts unchanged: the component gates still
need deafened/ASF conditions no golden combat has.
- SpellSpec.components (frozenset of V/S/M/F/DF, JSON key accepting the RAW
string notation "V, S, M/DF" or a token list; slash = alternative; unknown
token is a data error). Default empty = not encoded yet -> no failure.
- Combatant.armor_spell_failure (0-100 percent, authored on the JSON - the
model has no equipment) and arcane_caster (divine casters never incur ASF);
both load directly through load_combatant.
- _cast_spell: after the start-of-casting gates and before delivery, two
conditional d100 rolls (convention mirrors the concealment roll, <=):
deafened + V component spoils the spell on <=20 (SRD cast-a-spell.md);
arcane_caster + S component + armor_spell_failure > 0 fails the spell on
<= the ASF percent (SRD arcane-spells-and-armor.md). The slot stays spent
(mirrors the concentration fizzle). No condition -> no roll: the RNG
stream is bit-identical for every current combat (registry encodes no
components yet - full encoding is a follow-up task).
- Observation: spell summary gains "components" (sorted tuple), combatant
snapshot gains "armor_spell_failure"/"arcane_caster"; README documents
the keys and the two former out-of-scope bullets now describe the modeled
rule with its data-encoding follow-up.
- Data smoke test: lich arcane_caster true / ASF 0, oni arcane_caster false.
- tests/test_spell_failure.py: 21 tests (parsing, thresholds <=20 / <=ASF,
divine/no-ASF/no-component never roll, roll order, slot stays spent,
observation keys). Golden transcripts unchanged.
Adds the alignment field + axis flags, the AlignmentGate on spell damage
(Shard of Chaos, Spear of Purity) and the attacker-alignment-conditional
wards (Protection from Evil/Law + communal). Alignment leaves the
out-of-scope list; objects/inventory, skills, nonlethal and ASF are recorded
as the remaining out-of-scope mechanics. Gate 1829.
Add attacker-alignment modifier conditions (attacker_is_evil/lawful/good/
chaotic) evaluated against an optional attacker context threaded through
_stat_modifiers/_modifier_condition_holds, _attack_ac_mods (resolve_attack
AC path), resolve_save and _area_save (spell save paths), plus the fear
aura. Without a source in scope the condition is False, so existing rolls
are untouched (golden transcripts bit-identical).
Give the four protection_from_evil/law(+communal) JSONs their SRD buff:
+2 deflection to AC and +2 resistance on fort/ref/will, gated on the
attacker/caster alignment (SRD: bonuses apply 'against attacks or effects
created by evil creatures'); duration 1 min/level approximated as 10
rounds (engine convention). Descriptions document the unmodeled mental-
control suppression and communal duration-split clauses. Level/range/
save/SR verified against the KB.
DamageEffect gains an optional AlignmentGate (axis law_chaos/good_evil):
the immune side (chaotic/good) is wholly unaffected (no damage dice rolled),
neutral-on-axis targets take half damage with no condition (gate halving
stacks after the save halving), and the full side (lawful/evil) takes full
damage plus the gate condition on a failed save. Shard of Chaos maps the
SRD slowed to the engine staggered condition (the -1 penalty and half-speed
of slow are not modeled); the outsider-only 1d6/level variants are not
modeled (no outsider subtype). Non-gated damage spells are bit-identical.
Loader validates axis/side pairing with clear ValueErrors.
Adds the PF1e two-axis alignment code (LG/NG/CG/LN/N/CN/LE/NE/CE,
default N) to the Combatant model with a strict validator, derived
is_lawful/is_chaotic/is_good/is_evil properties, exposure in the
observation snapshot, and alignment values for all monsters and PCs
sourced from the bestiary KB and the Foundry sheets.
Guidance (new OneShotEffect), resistance (+1 resistance save buff, with the
cantrip-skip policy guard), virtue (temporary HP track) and stabilize
(auto-stabilization) are now modeled; spiritual_weapon and saving_finale
remain deferred with a precise engine reason. Gate 1734.
Rounds 1-3 closed the assumed residuals: run bounded by CON endurance + Run
feat, harmless spells forgo the save on willing targets, self-centered bursts
for bless/bane, Selective Channeling, the lich stat block aligned to the KB,
summon-stats disambiguation, healing_done exposure, multi-cell difficult-terrain
move pricing, and the grazed-caster-cell cone/line fix. The burst intersection
origin is deferred with a precise rationale. Rewrites the residual table to
drop the now-fixed and stale rows, leaving only the genuinely remaining items.
An independent pass confirmed the sampled High/Med fixes are genuinely
implemented, then found previously-undocumented divergences in the default
policy (spell combat without a slot, out-of-range heal wasting the turn,
wasted grit on a melee deed, fixed hex range), dice (wrong exception for
d1), the channel-heal trigger, and the foundry migration tool. All fixed;
corrects the stale 'goblin dies round 4' note (now round 5) and records the
Bless centering residual.
Records the 12 fixes of the final pass (engine Lows, LOS area templates,
runner/metrics/observation/CLI): dying creatures occupy their square,
blinded cannot charge/run/withdraw, cone/line aim and caster footprint,
square_size_ft validation, channel-heal no longer counted as damage dealt,
summon aggregation no longer KeyErrors, non-attack damage counted, README
examples refreshed, ACTION_RESULT_REASONS completed, zone-letter validation.
The remaining 35 spells were field-verified against the KB with zero
outcome divergences, closing the spell coverage gap.
"notes":"SRD Bestiary wolf (2d8+4 hp 13, AC 14/touch 12/flat 12, Fort +5, Will +1, bite +2 1d6+1 plus trip, Str 13 Dex 15 Con 15) + celestial template. Encoded from the template: DR 5/magic, energy resistance acid/cold/electricity 5, SR 6 (CR 1 + 5). Not encodable in this model: smite evil 2/day (+2 damage vs evil), the bite's free trip on hit (the engine has trip only as a standard-action maneuver, no auto-trip attack qualifier), and the weapon model has no magic property so no equipment bypasses the DR 5/magic (correct for mundane weapons). Summoning scaling (energy resistance +2 and SR +5 per summon spell level above 2nd) is not modeled; this sheet is the summon monster II baseline.",
"notes":"SRD Bestiary wolf (2d8+4 hp 13, AC 14/touch 12/flat 12, Fort +5, Will +1, bite +2 1d6+1 plus trip, Str 13 Dex 15 Con 15) + celestial template. Encoded from the template: DR 5/magic, energy resistance acid/cold/electricity 5, SR 6 (CR 1 + 5). Not encodable in this model: smite evil 2/day (+2 damage vs evil), the bite's free trip on hit (the engine has trip only as a standard-action maneuver, no auto-trip attack qualifier), and the weapon model has no magic property so no equipment bypasses the DR 5/magic (correct for mundane weapons). Summoning scaling (energy resistance +2 and SR +5 per summon spell level above 2nd) is not modeled; this sheet is the summon monster II baseline. Alignment inferred: SRD celestial template = any good; NG chosen for the good axis with neutral law axis",
"notes":"Stat block aligned to the KB lich entry (bestiary/lich.md): hp 111 (11d6+55 plus 15 false life), Str 10, Int 22, Init +2, fear aura 60 ft / DC 18. The spell list is the sim's own encounter-authored list (the KB's prepared list includes spells outside the data/spells registry); only the stat block was aligned.",
"notes":"Light-horse mount shared with Esha's sheet: hooves are 1d4 bludgeoning (SRD Bestiary horse, light - 2 hooves 1d4), natural attacks. No bite encoded: the SRD light horse has no bite attack and the Foundry sheet lists no mount bite (unlike Jeanne's sheet, which does).",
"notes":"Light-horse mount shared with Esha's sheet: hooves are 1d4 bludgeoning (SRD Bestiary horse, light - 2 hooves 1d4), natural attacks. No bite encoded: the SRD light horse has no bite attack and the Foundry sheet lists no mount bite (unlike Jeanne's sheet, which does). Alignment inferred: SRD animal (horse) is typically neutral; no alignment on the sheet",
"notes":"Traits : 'Born Damned' est un no-op documenté (+2 sauvegardes vs malédictions, descripteur 'curse' absent du moteur) ; 'Axe to Grind' est un no-op documenté (+1 dégâts vs ennemi menacé seulement par soi, condition non exprimable) ; 'Scaled Skin (electricity)' est encodé dans resistances (électricité 5, SRD : remplace Fiendish Resistance, donc pas de résistance froid/feu). 'Maw or Claw' : la fiche Foundry note la morsure « 1d4 piercing estimé » (item sans stats, estimation de l'auteur, non délibérée) ; le SRD (ARG, trait tieffelin daemon-spawn) accorde une morsure 1d6 à une créature Medium - l'encodage suit le SRD (1d6) et consigne l'écart avec la fiche.",
"name":"Attaques naturelles (Shifter Claws / formes fées)",
@@ -92,7 +99,7 @@
}
],
"source":"fiches_personnages/oni_sheet.json",
"notes":"Armes placeholder Foundry sans dégâts encodés omises : 'une Pioche' (non maîtrisée) et 'Weapon' (à remplacer par la vraie arme).",
"notes":"Armes placeholder Foundry sans dégâts encodés omises : 'une Pioche' (non maîtrisée) et 'Weapon' (à remplacer par la vraie arme). Traits : 'Exalted of the Society' accorde +1 canalisation d'énergie/jour (channel_energy 6 -> 7, fusionné par le chargeur) ; 'Extremely Fashionable' est un no-op documenté (+1 Bluff/Diplomacy/Intimidate en habits fins, état vestimentaire non modélisé) ; 'Bitter' (drawback) réduit de 1 les soins reçus d'un allié (comportement moteur).",
"notes":"Traits : 'Fortified Drinker' est un no-op documenté (+2 sauvegardes vs effets mentaux après alcool, état d'ébriété non modélisé) ; 'Lonely' (drawback) applique -2 Sense Motive et -2 sauvegardes vs charme (mind_affecting remplace charme).",
"features":[
"Weapon Focus (Pistol)",
"Rapid Reload (Pistol)",
@@ -202,5 +210,28 @@
"Delay Poison",
"Aid",
"Drunkard's Breath"
],
"carried_items":[
"Mess Kit",
"Gunsmith's Kit",
"Holy symbol (Wood)",
"Spell component pouch",
"Powder horn",
"Firearm Bullet",
"Alchemical Cartridge (Paper, Bullet Or Pellet)",
"Black Powder (Dose)",
"Matériaux pour munitions",
"Rope",
"Bouteille d'alcool (poste des bandits)",
"Clefs de la loge (4)",
"Black Powder (Dose) en stock",
"Firearm Bullet en stock",
"Alchemical Cartridge (Paper, Bullet Or Pellet) en stock",
"Alcool de poison d'araignée ogre",
"Badge des éclaireurs",
"Bouteille d'alcool fort (genre plus que du désinfectant)",
"description":"SRD spells/acid-arrow.md: ranged touch attack dealing 2d4 acid damage, plus another 2d4 acid damage at the start of each of the target's turns for 1 round + 1 round per three caster levels (maximum 6 additional rounds). Recurring damage is reduced by energy resistance/immunity, not by DR. No save."
"description":"SRD spells/acid-arrow.md: ranged touch attack dealing 2d4 acid damage, plus another 2d4 acid damage at the start of each of the target's turns for 1 round + 1 round per three caster levels (maximum 6 ADDITIONAL rounds at 18th level, i.e. 7 recurring rounds total). Recurring damage is reduced by energy resistance/immunity, not by DR. No save."
"description":"The subject gains a +1 morale bonus on attack rolls and on saving throws against fear effects (modeled as Will). Duration 10 min/level approximated as 10 rounds."
"description":"SRD spells/aid.md: the subject gains a +1 morale bonus on attack rolls and on saves against fear effects, plus temporary hit points equal to 1d8 + caster level (maximum 1d8+10). The save bonus is scoped to fear effects via the descriptor-conditioned modifiers (all three saves, replacing the previous unconditional Will-only approximation). The temporary hit points are deferred: CombatantState has no temp-HP track (same gap as the virtue cantrip). Duration 1 min/level approximated as 10 rounds."
"description":"SRD spells/bane.md: enemies within a 50-ft burst centered on the caster take a -1 penalty on attack rolls and a -1 penalty on saving throws against fear effects; Will negates. The -1 penalty is approximated with the shaken condition (-2 on attacks and saves), matching the existing Doom approximation. Duration 1 min/level approximated as 10 rounds."
"description":"SRD spells/bane.md: enemies within a 50-ft burst centered on the caster take a -1 penalty on attack rolls and a -1 penalty on saving throws against fear effects; Will negates. Modeled with the dedicated bane condition (untyped -1 on attacks, -1 on saves conditioned on the fear descriptor) - the previous shaken (-2) approximation overcounted the penalty. Duration 1 min/level approximated as 10 rounds."
"description":"The subject gains a +4 enhancement bonus to Constitution, granting +2 to Fortitude saves and +2 hit points per Hit Die (SRD bear's endurance; the hp modifier is informational — the engine has no per-HD hp pool). Will negates (harmless). Duration 10 min/level approximated as 60 rounds."
"description":"SRD spells/bear-s-endurance.md: the subject gains a +4 enhancement bonus to Constitution - +2 to Fortitude saves and +2 hit points per Hit Die - for 10 min./level. The engine has no ability-score track, so the +4 Con is encoded as its combat consequences: an unscaled +2 Fortitude enhancement, and a separate hp modifier carrying the per-HD bonus via flat scaling (+2 per level, capped at +40 for level 20; caster level stands in for the target's Hit Dice, and the hp target is informational - the engine has no per-HD hp pool). Will negates (harmless). Duration 10 min/level approximated as 60 rounds."
"description":"SRD spells/bleed.md: a living creature at -1 or fewer hit points begins dying, taking 1 point of damage per round; Will negates. Modeled as the bleeding condition for 1 round with bleed_damage = 1 (the engine's bleed tick deals it at the start of the target's turn)."
"description":"SRD spells/bleed.md: you target one living creature that has -1 or fewer hit points; it begins dying, taking 1 point of damage per round; Will negates. Gated on the target's hit points (requires_dying): a creature at 0 hp or above is not a valid target and the spell has no effect. Modeled as the bleeding condition for 1 round with bleed_damage = 1 (the engine's bleed tick deals it at the start of the target's turn)."
"description":"The charmed person regards the caster as a trusted friend (modeled as the fascinated condition; the charmed condition is not in the engine). Duration 1 hour/level approximated as 100 rounds. Will negates."
"description":"SRD spells/charm-person.md: range close (25 ft. + 5 ft./2 levels) - corrected from the previous medium approximation. The charmed person regards the caster as a trusted friend; the engine has no charmed condition, so it is modeled as fascinated (the subject takes no actions). Duration 1 hour/level approximated as 100 rounds. Will negates."
"description":"A cantrip: the subject gains a +1 competence bonus on a single attack roll, saving throw, or skill check, declared before the roll (duration 1 minute or until discharged). Will negates (harmless). Single-roll discharge is not modeled by the engine, so the spell has no mechanical effects here."
"effects":[
{
"type":"one_shot",
"track":"attack",
"value":1,
"bonus_type":"competence",
"duration_rounds":10
}
],
"description":"SRD spells/guidance.md: a cantrip - the subject gets a +1 competence bonus on a SINGLE attack roll, saving throw, or skill check, declared before the roll (duration 1 minute or until discharged). Modeled as a one-shot +1 competence modifier on the ATTACK track: the subject's next weapon attack roll (resolve_attack) discharges it and the bonus rides that roll only, not the crit-confirm roll; spell touch-delivery attacks roll on a separate path and do not consume it. The SRD 'attack OR save OR skill' choice has no declaration channel in the engine, so the attack-roll variant is encoded (the mechanism also carries save and skill tracks). 1 minute = 10 rounds, expiring at the start of the bearer's turns like any duration-tracked effect. The bonus folds into the resolve_modifiers pool, so it stacks with other types but not with a higher competence bonus on the same roll. default_policy never casts it (cantrips have no spell slot, and the one-shot type is not a buff/heal/offensive pick)."
"description":"A blast of intense cold coats a 5-ft-radius burst in ice: creatures in the area take 1d6 points of cold damage (Reflex half) and fall prone (no prone on a successful save). The lingering slippery surface (half speed, Acrobatics checks) is not modeled."
"description":"SRD spells/ice-slick.md: a blast of intense cold coats a 5-ft-radius burst in ice: creatures in the area take 1d6 points of cold damage + 1 point per caster level (maximum +10) and fall prone; a successful Reflex save halves the damage and avoids falling prone. The lingering slippery surface (half speed, Acrobatics checks) is not modeled."
"description":"Illusory doubles (1d4 + 1 per 3 caster levels, max 8) occupy your space, making attacks risk striking a figment instead. Duration 1 min/level approximated as 10 rounds. Simplification: the image-destruction mechanic is approximated as the blurred condition (20% miss chance); the real spell can deflect attacks entirely until the images are used up."
"description":"SRD spells/mirror-image.md: 1d4 illusory doubles plus one per three caster levels (maximum eight total) inhabit your space. The image count is rolled at cast time and folded into the engine's concealment machinery as a dynamic miss chance of images/(images+1) on the blurred carrier condition (an attack that would hit strikes a figment instead). Duration 1 min./level scaled as 10 rounds per caster level. Simplification: the figment-destruction track (a hit or a near miss by 5 or less destroys one image; area and no-attack-roll effects ignore them) is not modeled - the miss chance persists for the whole duration."
"description":"The subject gains a +4 enhancement bonus to Wisdom for 1 min/level, adding the usual benefit to Wisdom-based skills, Will saves and spell DCs. Will negates (harmless). Ability-score increases are not modeled by the engine, so the spell has no mechanical effects here."
"description":"SRD spells/owl-s-wisdom.md: the subject gains a +4 enhancement bonus to Wisdom for 1 min./level. The engine has no ability-score track, so the +4 Wis is encoded as its combat-visible consequence: a +2 enhancement bonus on Will saves (half of +4, as a score modifier feeds saves at half). Wisdom-based skills and spell DCs are not modeled. Will negates (harmless). Duration 1 min/level approximated as 10 rounds."
"description":"A barrier wards the subject from evil for 1 min/level: +2 deflection bonus to AC and +2 resistance bonus on saves against attacks and effects created by evil creatures, renewed saves and suppression of mental control, and lawful summoned evil creatures cannot touch it. Will negates (harmless). The engine has no alignment system, so alignment-restricted bonuses cannot be modeled and the spell has no mechanical effects here."
"description":"SRD spells/protection-from-evil.md: a barrier wards the subject from evil for 1 min./level. The subject gains a +2 deflection bonus to AC and a +2 resistance bonus on saves - the exact SRD wording: \"+2 deflection bonus to AC and a +2 resistance bonus on saves against attacks or effects created by evil creatures\" - so both bonuses are gated on the attacker/caster alignment (condition attacker_is_evil) and never apply against a creature that is not evil. Will negates (harmless), spell resistance no. Duration 1 min/level approximated as 10 rounds (engine convention for 1-minute durations). Not modeled: the SRD's second effect (immediate renewed saving throw with a +2 morale bonus against mental-control spells and effects, suppression for the duration, immunity to new attempts) - the engine has no mental-control tracking to suppress; and the third effect (evil summoned creatures cannot touch the warded creature) - the engine has no summoned-subtype contact restriction."
"description":"Functions like protection from evil (+2 deflection to AC and +2 resistance on saves vs evil creatures, protection from mental control), except the 1 min/level duration is divided in 1-minute intervals among the touched creatures. Will negates (harmless). Alignment-restricted bonuses and duration-sharing are not modeled, so the spell has no mechanical effects here."
"description":"SRD spells/protection-from-evil-communal.md: functions like protection from evil - each touched creature gains the +2 deflection bonus to AC and +2 resistance bonus on saves against attacks or effects created by evil creatures (gated on the attacker/caster alignment via condition attacker_is_evil, never applying against a non-evil source). Will negates (harmless), spell resistance no. Not modeled: the SRD duration split - \"you divide the duration in 1-minute intervals among the creatures touched\" - the engine gives every touched creature the same fixed duration as protection from evil (1 min/level approximated as 10 rounds each) instead of draining one shared pool; and the mental-control suppression clause of the base spell (see protection_from_evil.json)."
"description":"A barrier wards the subject from lawful creatures for 1 min/level: +2 deflection bonus to AC and +2 resistance bonus on saves against attacks and effects created by lawful creatures, plus renewed saves against mental control by lawful creatures. Will negates (harmless). The engine has no alignment system, so alignment-restricted bonuses cannot be modeled and the spell has no mechanical effects here."
"description":"SRD spells/protection-from-law.md: functions like protection from evil, except the deflection and resistance bonuses apply against lawful creatures - \"+2 deflection bonus to AC and a +2 resistance bonus on saves against attacks or effects created by lawful creatures\" - so both bonuses are gated on the attacker/caster alignment (condition attacker_is_lawful) and never apply against a creature that is not lawful. Will negates (harmless), spell resistance no. Duration 1 min/level approximated as 10 rounds (engine convention for 1-minute durations). Not modeled: the renewed saving throw against mental control by lawful creatures and the lawful summoned creature contact restriction (same gaps as protection from evil)."
"description":"Functions like protection from law (+2 deflection to AC and +2 resistance on saves vs lawful creatures, protection from mental control), except the 1 min/level duration is divided in 1-minute intervals among the touched creatures. Will negates (harmless). Alignment-restricted bonuses and duration-sharing are not modeled, so the spell has no mechanical effects here."
"description":"SRD spells/protection-from-law-communal.md: functions like protection from law - each touched creature gains the +2 deflection bonus to AC and +2 resistance bonus on saves against attacks or effects created by lawful creatures (gated on the attacker/caster alignment via condition attacker_is_lawful, never applying against a non-lawful source). Will negates (harmless), spell resistance no. Not modeled: the SRD duration split - \"you divide the duration in 1-minute intervals among the creatures touched\" - the engine gives every touched creature the same fixed duration as protection from law (1 min/level approximated as 10 rounds each) instead of draining one shared pool; and the mental-control suppression clause of the base spell (see protection_from_law.json)."
"description":"A cantrip: the subject gains a +1 resistance bonus on saving throws for 1 minute. Will negates (harmless). Simplified: the engine has no resistance bonus type and no self-buff encoding that preserves the pinned default-policy casting order, so no mechanical effect is applied."
"description":"SRD spells/resistance.md: a cantrip - the subject gains a +1 resistance bonus on saving throws for 1 minute. Will negates (harmless). Modeled as a BuffEffect granting +1 resistance to fort/ref/will for 10 rounds (1 minute); the resistance bonus type does not stack with itself, so a second Resistance (or a cloak of resistance) keeps the higher value. default_policy never casts it: cantrips have no spell slot, and _find_buff_spell skips spells the caster has no slot for, so Resistance cannot preempt a leveled buff (Divine Favor) in the known-spell order."
"description":"Immediate action, requires an active bardic performance: when a creature affected by your performance fails a saving throw, end the performance to let it immediately reroll the failed save and take the new result. Will negates (harmless). No combat effect encoded: conditional save rerolls tied to bardicperformance are not modeled by the engine."
"description":"SRD spells/saving-finale.md: immediate action, requires an active bardic performance - when a creature affected by your bardic performance fails a saving throw, end the performance to let it immediately reroll the failed save and take the new result. Will negates (harmless). Deferred: rerolling an already-resolved save needs a post-save reaction hook plus a bardic-performance resource gate (the engine's saves are one-shot resolutions with no reroll window), a cross-cutting redesign. No combat effect encoded; the immediate casting time is kept for data fidelity."
"description":"SRD spells/shard-of-chaos.md: evocation [chaotic], level 2 (cleric/oracle/warpriest), range close, dart-shaped projectile of chaotic energy delivered as a ranged touch attack, Will partial, spell resistance yes. A lawful creature struck takes 1d8 points of damage per two caster levels (maximum 5d8); a successful Will save halves the damage. Simplified: the engine has no alignment track, so the SRD alignment gate (no effect on chaotic creatures, half damage to neutrals, the lawful-outsider 1d6/level + slowed variant) cannot be represented - the base 1d8-per-two-levels damage (engine floor: caster_level // 2 dice) is modeled against every target as force damage (no energy type models chaotic energy), Will half."
"description":"SRD spells/shard-of-chaos.md: evocation [chaotic], level 2 (cleric/oracle/warpriest), range close, dart-shaped projectile of chaotic energy delivered as a ranged touch attack, Will partial, spell resistance yes. A lawful creature struck takes 1d8 points of damage per two caster levels (maximum 5d8) and, on a failed Will save, is slowed for 1 round; a successful Will save halves the damage and negates the slow. The bolt has no effect on chaotic creatures (alignment_gate immune) and deals only half damage to creatures neither lawful nor chaotic, without the slow (alignment_gate neutral halving). Simplifications: the SRD's slowed condition is mapped to the engine's staggered condition (the SRD slow entry states a slowed creature is staggered; the -1 penalty and half-speed parts of slow are not modeled); the lawful-outsider variant (1d6 per casterlevel, maximum 10d6) is not modeled (the engine has no outsider subtype); force damage stands in for chaotic energy (no energy type models it)."
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
ctan